Virtual Reality by Mobile Phone: Improving Child Pedestrian Safety Pedestrian injuries are a leading cause of pediatric injury. Effective, practical, and cost-efficient empirically- based and theoretically-driven behavioral interventions to teach young children street-crossing skills are needed. Our laboratory has explored virtual reality (VR) as a means to teach child pedestrian safety skills for several reasons, including: (a) it offers children repeated unsupervised practice in pedestrian crossings without risk of injury, (b) it provides automated feedback to children on crossing success or failure, (c) it can be tailored to child skill levels: (d) it offers an appealing and fun training environment, and (e) most recently given technological advances, it offers potential for broad dissemination using mobile phone technology. Previous work indicated VR was an effective pedestrian safety training tool. The present proposal extends previous findings in two critical ways. First, it will evaluate delivery of a pedestrian VR using a mobile phone and the Google Cardboard platform, technology released in 2014 that enables a standard mobile phone to be used as an immersive virtual reality delivery system. Second, it will overcome the limitation of previous research suggesting children learned some pedestrian skills after six 30-minute training sessions in a VR but did not yet master adult-level pedestrian skills. We will implement a randomized non-inferiority trial with two equal-sized groups of children ages 7-8 (total N = 498). All participants will complete baseline, post-intervention, and 6- month follow-up assessments of pedestrian safety and up to twenty-five 30-minute pedestrian safety training trials until they reach adult levels of functioning. Half the children will be randomly assigned to train in Google Cardboard and the other half in an existing semi-immersive laboratory VR. Primary outcomes will be assessed using ANCOVA models.